Add agt_Latn shard 0: 3592 sentences, 160 paragraphs, 4157 words, 190702 ngrams

README.md

@@ -531,22 +531,6 @@ configs:
     data_files:
       - split: train
         path: data/ngrams/cpc_Latn/*.parquet
-  - config_name: sentences-qvc_Latn
-    data_files:
-      - split: train
-        path: data/sentences/qvc_Latn/*.parquet
-  - config_name: paragraphs-qvc_Latn
-    data_files:
-      - split: train
-        path: data/paragraphs/qvc_Latn/*.parquet
-  - config_name: words-qvc_Latn
-    data_files:
-      - split: train
-        path: data/words/qvc_Latn/*.parquet
-  - config_name: ngrams-qvc_Latn
-    data_files:
-      - split: train
-        path: data/ngrams/qvc_Latn/*.parquet
   - config_name: sentences-xsr_Deva
     data_files:
       - split: train
@@ -563,6 +547,22 @@ configs:
     data_files:
       - split: train
         path: data/ngrams/xsr_Deva/*.parquet
+  - config_name: sentences-qvc_Latn
+    data_files:
+      - split: train
+        path: data/sentences/qvc_Latn/*.parquet
+  - config_name: paragraphs-qvc_Latn
+    data_files:
+      - split: train
+        path: data/paragraphs/qvc_Latn/*.parquet
+  - config_name: words-qvc_Latn
+    data_files:
+      - split: train
+        path: data/words/qvc_Latn/*.parquet
+  - config_name: ngrams-qvc_Latn
+    data_files:
+      - split: train
+        path: data/ngrams/qvc_Latn/*.parquet
   - config_name: sentences-waj_Latn
     data_files:
       - split: train
@@ -1000,8 +1000,8 @@ configs:
 # FineWeb-2 NLP
 
 > Every sentence, paragraph, word, and n-gram from [FineWeb-2](https://huggingface.co/datasets/HuggingFaceFW/fineweb-2),
-> split with language-aware segmentation — **1,130,777 sentences** across **245 languages**,
-> from **19,704 source documents** — updated continuously.
+> split with language-aware segmentation — **1,134,369 sentences** across **246 languages**,
+> from **19,864 source documents** — updated continuously.
 
 ## Table of Contents
 
@@ -1031,10 +1031,10 @@ useful units, ready for researchers and practitioners:
 
 | Type | Rows | What you get |
 |------|------|-------------|
-| **sentences** | 1,130,777 | One row per sentence, with source document ID, URL, and position index |
-| **paragraphs** | 20,169 | One row per paragraph, with sentence count per paragraph |
-| **words** | 1,648,619 | Per-shard word frequency and document frequency tables |
-| **ngrams** | 46,401,003 | Per-shard bigram through 5-gram frequency tables |
+| **sentences** | 1,134,369 | One row per sentence, with source document ID, URL, and position index |
+| **paragraphs** | 20,329 | One row per paragraph, with sentence count per paragraph |
+| **words** | 1,652,776 | Per-shard word frequency and document frequency tables |
+| **ngrams** | 46,591,705 | Per-shard bigram through 5-gram frequency tables |
 
 Every row traces back to its source document via `doc_id` and `doc_url`, so you can always
 navigate from a sentence or word back to the original web page.
@@ -1177,23 +1177,23 @@ GROUP BY lang ORDER BY sentences DESC LIMIT 30;
 swg_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 301
 liv_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 227
 got_Goth       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 162
-eko_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 161
 sab_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 161
+eko_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 161
 zyp_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 161
-guo_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
-bim_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
+agt_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 160
 xmv_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
+guo_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
 iws_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
 bsq_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
+bim_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 158
 nhg_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 157
 cao_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 153
-mak_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 153
 caa_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 153
 llg_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 153
+mak_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 153
 mbl_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 152
-zab_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 151
 mcd_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 151
-dng_Cyrl       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 149
+zab_Latn       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 151
 ```
 
 <details>
@@ -1348,16 +1348,16 @@ print(words_df)
 
 | Metric | Value |
 |--------|-------|
-| **Total sentences** | **1,130,777** |
-| **Total paragraphs** | **20,169** |
-| **Total word entries** (per-shard) | **1,648,619** |
-| **Total n-gram entries** (per-shard) | **46,401,003** |
-| **Languages processed** | **245** |
-| **Source documents** | **19,704** |
-| **Total parquet size** | **537.0 MB** |
-| Avg sentences per document | 57.4 |
+| **Total sentences** | **1,134,369** |
+| **Total paragraphs** | **20,329** |
+| **Total word entries** (per-shard) | **1,652,776** |
+| **Total n-gram entries** (per-shard) | **46,591,705** |
+| **Languages processed** | **246** |
+| **Source documents** | **19,864** |
+| **Total parquet size** | **538.8 MB** |
+| Avg sentences per document | 57.1 |
 | Avg paragraphs per document | 1.0 |
-| Avg sentences per paragraph | 56.1 |
+| Avg sentences per paragraph | 55.8 |
 
 ### Per-language breakdown
 
@@ -1394,8 +1394,8 @@ print(words_df)
 | 29 | `bim_Latn` | 8,508 | 158 | 7,224 | 455,956 | 158 | 1 | 4.3 MB |
 | 30 | `mav_Latn` | 8,495 | 91 | 5,784 | 344,782 | 91 | 1 | 3.6 MB |
 | 31 | `cpc_Latn` | 8,404 | 126 | 15,100 | 207,140 | 126 | 1 | 3.1 MB |
-| 32 | `qvc_Latn` | 8,402 | 116 | 14,058 | 249,877 | 116 | 1 | 3.3 MB |
-| 33 | `xsr_Deva` | 8,402 | 141 | 1,723 | 281,923 | 141 | 1 | 2.8 MB |
+| 32 | `xsr_Deva` | 8,402 | 141 | 1,723 | 281,923 | 141 | 1 | 2.8 MB |
+| 33 | `qvc_Latn` | 8,402 | 116 | 14,058 | 249,877 | 116 | 1 | 3.3 MB |
 | 34 | `waj_Latn` | 8,257 | 113 | 7,789 | 309,979 | 113 | 1 | 3.6 MB |
 | 35 | `eko_Latn` | 7,963 | 161 | 12,444 | 336,501 | 161 | 1 | 4.0 MB |
 | 36 | `kwf_Latn` | 7,960 | 143 | 4,567 | 375,401 | 123 | 1 | 3.7 MB |

data/ngrams/agt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf7b9bed2bdbd503a164c25b7aebe9cbd7b400d53249f735d35d3024ca5318b9
+size 1539974

data/paragraphs/agt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84bc3582922d710aae2b5cd7a0e9314e0308cca340f37511c71879cb18f8790f
+size 181850

data/sentences/agt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae889a8f06e5d47c661fd6892c216233b7f312fe834975a65272751b4cefbd7d
+size 128196

data/words/agt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:757e0a759db199b7209db42c7efe21dc90c5e170edad2481034e4548dcd8cae6
+size 20989

stats.csv

@@ -244,3 +244,4 @@ zyp_Latn,zyp,0,159,7183,161,9889,458078,4834159,2026-04-14T20:56:06Z
 liv_Latn,liv,0,159,5892,227,16639,229479,3251278,2026-04-14T20:56:25Z
 sab_Latn,sab,0,161,8663,161,3372,482580,4855264,2026-04-14T20:56:47Z
 eko_Latn,eko,0,161,7963,161,12444,336501,4232517,2026-04-14T20:57:05Z
+agt_Latn,agt,0,160,3592,160,4157,190702,1871009,2026-04-14T20:57:26Z